14 Oti the Laws of Terrestrial Magnetism 



with some very irregular inflexions. According to the most re- 

 cent observations, there exists at present a line of no variation 

 in the Atlantic Ocean between tlic old and the new world. It 

 cuts the meridian of Paris, near the south latitude of 65°; from 

 this it proceeds in a NNW direction towards about 35° of west 

 longitude, whence it reaches as high up as the coasts of Para- 

 guay; then, after going on nearly north and south, it runs along 

 the coasts of Brasil, and thus proceeds to the latitude of Ca- 

 yenne : but then turning suddenly to the NW, it directs its 

 course towards the United States, and thence towards the other 

 northern parts of the continent of America, which it traverses in 

 the same direction. : 



The position of this line of no variation is not fixed on the 

 globe ; at least it has not been so during the last century and a 

 half; but has moved considerably from the east towards the 

 west. It passed through London in 1657, and through Paris in 

 1664*. So that, according to its present direction, it has de- 

 scribed about SO" of longitude in 150 years along this parallel. But 

 there is no doubt that this alteration is not uniform in its mo- 

 tion : it is, indeed, very unequal in different parallels ; at Ja- 

 maica, for example, the variation has not experienced any sen- 

 sible change for 140 years. In general, according to the pre- 

 sent slowness of this motion, we are not certain that it is always 

 progressive, nor that it proceeds in any given direction. These 

 are things that time alone can bring to light. 



Some very accurate observations of the dip, made at different 

 times by the Honourable Mr. Cavendish and Mr. Gilpin in Lon- 

 don, have proved that this element is likewise variable, although 

 much less so than the variation. In 1775, the dip in London 

 was 72' 30': in 1805, it was 70*' 21'. This has also been con- 

 firmed in France by the experiments of M. Humboldt. 



There exists another band having no variation nearly oppo- 

 site the preceding one : this runs in a continued direction to the 

 NW, takes its commencement in the great South Sea, cuts the 

 western point of New Holland, traverses the Indian Sea, enters 

 the continent of Asia at Cape Comorin, atiil thence passing 

 across Persia and Siberia westward, rises up towards Lapland. 

 But what is very remarkaljie, this line bifurcates near the great 

 archipelago of Asia, and gives rise to another branch, which, 

 directing itself" almost wholly from south to north, passes this 

 archipelago, crosses China, and returns into itself again in the 

 eastern part of Siberia. The existence of this branch, and it& 

 separation from the preceding one, are clearly pointed out by 

 observations made in the Chinese seas. But I am enabled to 



* See Walker's Treatise oii Magnetism, p. 197. 



present 



